Diurnal Cycle of Clouds and Precipitation at the ARM SGP Site

Millimeter Wavelength Cloud Radar (MMCR) data from Dec. 1996 to Dec. 2010, collected at the U. S. Department of Energy Atmospheric Radiation Measurement (ARM) program site in the U.S. Southern Great Plains (SGP), are categorized into clouds (-40dBZe≤reflectivity<-10dBZe), drizzle and light precipitation (-10dBZe≤reflectivity<10dBZe), and heavy precipitation (reflectivity≥10dBZe), and are used to examine the diurnal cycle of hydrometeors occurrence.   The same criteria are implemented for the observation-equivalent reflectivity calculated by feeding outputs from a Multiscale Modeling Framework (MMF) climate model simulation into a radar simulator.  The MMF model consists of the NCAR Community Atmosphere Model (CAM) with conventional cloud parameterizations replaced by a cloud resolved model (CRM). Observational and simulated radar reflectivity are compared and further sorted into different atmospheric states identified by Evans (2014).  Evans used a neutral network to take ERA-Interim state variables (i.e. horizontal winds, relative humidity, temperature at seven predetermined pressure level and surface pressure) on an 8×8 grid with 1.5º×1.5º spatial resolution centered on the SGP site and found twenty-one atmospheric states which represent specific synoptic conditions.  We use these states to study the differences in the diurnal cycle between observations and simulations.  Differences in the (mean) annual diurnal cycle between the observations and model are decomposed into errors in the daily mean, errors in the diurnal variation in each state, and errors due to difference in the frequency of occurrence of atmospheric states between ERA and the MMF. The magnitude of various error sources is assessed.